Modelling the spatio-temporal dynamics of youth unemployment in Africa: A Bayesian approach

Youth unemployment has remained one of the major social and economic problems facing many African countries despite persistent efforts by governments and other stakeholders to mitigate the problem. We deployed a two-parameter beta geospatial model within a distributional regression framework that enables us to link covariates to the mean and variance parameters of the response distribution, to study the spatio-temporal patterns of unemployment combining multiple African countries, while accounting for the impact of temperature and rainfall as climatic variables. We also examined the pattern of relationship between unemployment and gross domestic product of each of the countries over the study period. The data for the study was sourced from the World Bank Development Indicator database covering 1991 to 2020 while Bayesian inference was based on Markov chain Monte Carlo simulation. The findings reveal rising unemployment rates with increase in temperature and rainfall, and that in most cases, neighbouring African countries show similar unemployment pattern over the study period.